About

Mingyu Chen (Rayner) Mingyu Chen (Rayner)

Hi, I’m Mingyu Chen — Rayner online.

I’m the Co-Founder and VP of Engineering at VeloDB, and the PMC Chair of Apache Doris — an open-source real-time analytics and search database for the AI era.

Background

Twelve years building data systems. Before co-founding VeloDB, I was a Senior R&D Engineer at Baidu, where I worked on the distributed analytics infrastructure that would later become Apache Doris. I’ve been contributing to Doris since its early days — from the first open-source release to the project it is today.

These days my focus is on the architecture decisions that shape where Doris goes next: lakehouse integrations (Iceberg, Paimon, Hudi), semi-structured data (Variant), and the seam between analytics engines and open table formats.

About this blog

This is where I lay out the longer arguments that don’t fit in a GitHub thread, a Slack message, or a customer call. Topics tend to cluster around:

  • Lakehouse architecture — open table formats, how query engines integrate with them, where the hard problems actually live.
  • Open-source database internals — storage, execution, query optimization, and the tradeoffs behind specific design choices.
  • Data systems in the AI era — semi-structured workloads, observability for agent pipelines, and the ways analytics engines adapt to workloads they weren’t designed for.
  • Occasional reflections — on shepherding an open-source project and building a company around it.

Posts are in English. If something here matters to you, reach out — I usually enjoy the follow-up conversations more than the posts themselves.

Find me elsewhere